Visual customer segmentation

Customer segmentation is becoming one of the top priorities for digital marketers today. There are many methods that can be used. Let me describe one that isn’t based on an exact science, but is very effective.

I call it “visual segmentation”. Visual because you must visually assess you customer distribution by the given dimension to determine the segments.

The chart below shows an example of a customer portfolio. You can see the count of customers displayed by age. Each bar represents one age group and its size corresponds to the number of customers in that age group. The bars have been further split by customers’s gender, but this is not important for the thing I want to explain.

Visual segmentation on a histogram

 

Let’s look at the chart and try to find “valleys” among the bars. To avoid coincidences and to strengthen our assumption let’s consider only a valley where its lowest bar has at least two neighbouring bars that are higher and higher. This indicates segments of people. I can see two valleys that meet the criteria – with bottoms at age 30 and 56. I can interpret this to mean that there is a segment of young people up to 30 years old, middle-aged people from 30 to 56, and a third segment comprising people older than 56 years.

As a marketer I could then customize my email template’s design and phrasing so that the messages better resonate with these three age groups. By doing that, I could expect increased email open and click rates and corresponding conversions.

It depends case-by-case which segments you discover in your data. Check for strong patterns and other indications that support the existence of such segments. When you look at the distribution too much, your mind will likely create false assumptions.

What criteria do you use for customer segmentation? Please share your experience.

No Comments Yet

Leave a Reply

Your email address will not be published. Required fields are marked *

Get 5 FREE case studies!

Your information will never be shared with any third party.